Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2605.00462 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911638408396800 |
|---|---|
| author | Rosciszewski, P. Krzywaniak, A. Iserte, S. Rojek, K. Gepner, P. |
| author_facet | Rosciszewski, P. Krzywaniak, A. Iserte, S. Rojek, K. Gepner, P. |
| contents | Intelligence Processing Units (IPU) have proven useful for many AI applications. In this paper, we evaluate them within the emerging field of \emph{AI for simulation}, where traditional numerical simulations are supported by artificial intelligence approaches. We focus specifically on a program for training machine learning models supporting a \emph{computational fluid dynamics} application. We use custom TensorFlow provided by the Poplar SDK to adapt the program for the IPU-POD16 platform and investigate its ease of use and performance scalability. Training a model on data from OpenFOAM simulations allows us to get accurate simulation state predictions in test time. We show how to utilize the \emph{popdist} library to overcome a performance bottleneck in feeding training data to the IPU on the host side, achieving up to 34\% speedup. Due to communication overheads, using data parallelism to utilize two IPUs instead of one does not improve the throughput. However, once the intra-IPU costs have been paid, the hardware capabilities for inter-IPU communication allow for good scalability. Increasing the number of IPUs from 2 to 16 improves the throughput from 560.8 to 2805.8 samples/s. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_00462 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Adaptation of AI-accelerated CFD Simulations to the IPU platform Rosciszewski, P. Krzywaniak, A. Iserte, S. Rojek, K. Gepner, P. Distributed, Parallel, and Cluster Computing Artificial Intelligence Intelligence Processing Units (IPU) have proven useful for many AI applications. In this paper, we evaluate them within the emerging field of \emph{AI for simulation}, where traditional numerical simulations are supported by artificial intelligence approaches. We focus specifically on a program for training machine learning models supporting a \emph{computational fluid dynamics} application. We use custom TensorFlow provided by the Poplar SDK to adapt the program for the IPU-POD16 platform and investigate its ease of use and performance scalability. Training a model on data from OpenFOAM simulations allows us to get accurate simulation state predictions in test time. We show how to utilize the \emph{popdist} library to overcome a performance bottleneck in feeding training data to the IPU on the host side, achieving up to 34\% speedup. Due to communication overheads, using data parallelism to utilize two IPUs instead of one does not improve the throughput. However, once the intra-IPU costs have been paid, the hardware capabilities for inter-IPU communication allow for good scalability. Increasing the number of IPUs from 2 to 16 improves the throughput from 560.8 to 2805.8 samples/s. |
| title | Adaptation of AI-accelerated CFD Simulations to the IPU platform |
| topic | Distributed, Parallel, and Cluster Computing Artificial Intelligence |
| url | https://arxiv.org/abs/2605.00462 |